WT and overture data corruption?

is this possible?

[stargeek]

a term "widgets in uganda" is in the short term top 1000 in wt, and shows a bunch of traffic to it, overture's free keyword tool also show it having a good deal of traffic, but some obviously corrupt data is also in WT, ie: "wigetsinuganda" and things like that, where it is obvious that a bunch of real people are not actually searching for that term. I have a page ranked for this term and it appears that its not getting very much traffic at all, is it possible that WT *and* overture's data is skewed, by rank checkers or something like that?

is there any source of unassailable keyword data?

[keywordguru]

It is very likely that a portion may be automated queries while others may be legit.

Depending on your rank, it may also determine how deep people actually go to read up on a topic. I have seen some keywords hold their value while others do not. I would put some good tracking in place to see if any users go from that page to others and see if anything converts before investing tons more time into it.

Hope that helps
KG

[stargeek]

I'm on the first page of the serp for this term, and I'm getting next to no traffic at all.

both WT and overture show a ton of traffic.

I'm aware that overture's data can be corrupted, but WT uses metacrawler and dogpile, which i thought was not very often used for rank checking.

what would be the benefit of someone deliberatly inflating these numbers?

[thevoodoo]

WT and OV are pretty much the only major sources for keyword research next to smaller lists form some less popuar search engines that you could probably purchase.

however, choosing the right keywords can nto be done by just using the data from these two sources, at least not just search count data.

What you can do however, is compare the search count, overture top bids, and keyword competitiveness for each keyword which then will go you a much better idea s to whether the search coutn was fabricated or real.

For example high search count, low overture bid will mean two things: 1. not a profitable keyword as far as financial gains are conerned (probably free widget); or 2. a fabricated search count.

[edited by: mona at 8:25 pm (utc) on Oct. 13, 2005]
[edit reason] no urls, please - thx! [/edit]

[stargeek]

keyword discovery is also decent, but the data seems unususally high for this term on there also.